Set to Idle problem & Data Filtering GA4

[Selçuk Kocabıyık (Ingage)]

Hello Team,

 

Our propensity prediction pipeline is keep being set to idle on scheduled triggers. Warnings we recieve on the pipeline is only on Events to GA4 - Worker Class: BQToMeasurementProtocolGA4 as it is shown on the screenshot. After this job we do not find any success log on the latter jobs while prior job logs show both set to idle and success messages. There are too many logs being shown on the list due to, I assume, some error iterations.

Hence, we are not sure if it is working properly or not, is there any solution/explanation we can consider?

 

Also another question; is it possible to use GA4 testing data filtering by adding parameter to measurement protocol json content, so that while audiences will being updated no report will contain any bqml event or session data?

 

 

 

Thanks in advance,

 

-- 

Selçuk Kocabıyık Business Intelligence Manager

Ünalan Mahallesi, Ayazma Caddesi Çamlıca İş Merkezi A Blok No: 131-1, İç Kapı No:8 Üsküdar, İstanbul /Türkiye

www.ingage.media

 

---

Bu e-posta mesajı kişiye özel olup, gizli bilgiler içeriyor olabilir. Eğer bu e-posta mesajı size yanlışlıkla ulaşmışsa e-posta mesajını kullanıcıya hemen geri gönderiniz ve mesaj kutunuzdan siliniz. Bu e-posta mesajı hiçbir şekilde herhangi bir amaç için çoğaltılamaz, yayınlanamaz ve para karşılığı satılamaz. Yollayıcı bu e-posta mesajının – virüs koruma sistemleri ile denetleniyor olsa bile – virüs içermediğini garanti etmez ve meydana gelebilecek zararlardan doğacak hiçbir sorumluluğu kabul etmez.

Rekabet Kanunu'nun lafzına ve ruhuna uyum, Ingage'in hukuka uygun davranma politikasının önemli bir parçasıdır. İşbu elektronik postada; tarafınıza Rekabet Kanunu'nu ihlal edebilecek bir rekabete duyarlı bilgi iletildiği ve/veya talep edildiği yahut Rekabet Kanunu'nu ihlal edebilecek diğer herhangi bir hususa yer verildiği kanaatindeyseniz lütfen gecikmeksizin Ingage Dijital Pazarlama Hizmetleri A.Ş. Hukuk Departmanı ile iletişime geçiniz.

This message may contain confidential and/or privileged information. If you have received this message in error, please notify us immediately by responding to this e-mail and then delete it from your system. Thank you for your cooperation.

Compliance with both the letter and spirit of the Act on the Protection of Competition is an important part of Ingage's policy of legality. Please contact with Law Department of Ingage immediately if you have received and/or have been asked for a competitively sensitive information by this e-mail or if the e-mail includes any subject which may constitute a breach of competition legislation.

Sınıflandırma: INGAGE İÇİ / Classification: INTERNAL

[Instant BQML and Vertex Users]

Hello,

Thank you for reaching out to us about the issues you're experiencing with your propensity prediction pipeline and your question about GA4 testing data filtering.

Pipeline Idle Status:

I understand you're noticing that on certain runs your prediction pipeline is being set to idle. This is a known issue with pipeline execution. For certain runs, we see that there might be a duplicate finish request. When this happens, we set the pipeline to idle and await the next run.

The prediction pipeline is set to run every 6 hours and includes logic to handle any duplicate scoring within a 24-hour period, so the next run should score your users appropriately. This should not happen on every run, but we do notice it from time to time, and we are working to figure out a long-term fix. Rest assured, using the Instant BQML pipelines with the default cron should score your users comprehensively.

It may also be worthwhile to ensure you're running on the most updated version of CRMint by upgrading your application by running: 

• bash <(curl -Ls https://raw.githubusercontent.com/instant-bqml/crmint/master/scripts/install.sh) master --bundle

GA4 Testing Data Filtering:

Regarding your question about removing any reporting implications from your Instant BQML pipelines, you can remove the "Events to GA4" job from your prediction pipeline, which sends the event measurement protocol to your GA4 property. The "Percentiles to GA4" job sends your propensity score with the user property, which your audiences are defined based on, so that measurement protocol payload will increase your audience list sizes without impacting reporting.

To remove the "Events to GA4" job, you will likely need to:

1. Disable the "Run on Schedule" toggle for your Prediction Pipeline.
2. Click the 3 vertical dots next to the "Events to GA4" job and select "Remove."
3. Click into the "Percentiles to GA4" job, set the Job Dependencies to "Define User Property" and condition "success," save the job, and then re-enable your pipeline. It should be one linear pipeline at that point.
   1. See attached screenshots

Please let us know if you have any further questions or concerns.